As storage clusters and other storage systems grow in capacity, metadata becomes ever larger and evermore important to handle and store efficiently. Defining an address space and allocating memory for storage of all possible combinations within the address space is inefficient and wasteful when that address space is not fully occupied. Some schemas and storage systems have relatively long names in metadata, and storage of large amounts of such metadata with long names takes a large amount of storage space. There is a need for efficient name coding in storage systems, to efficiently use available memory space and to allow greater storage capacity for a given amount of memory, so that available memory space is not taken up by allocated but sparsely or not fully occupied data structures and by large amounts of metadata with long names. It is within this context that present embodiments arise.